Cumulative data on S65 engine failures.

[regular guy]

Originally Posted by SenorFunkyPants

The only way forward is for an Admin (or mod is they have access) to interrogate the M3post database for: Drives M3.
Copy all the entries and then manually check the field for E9x owners.
I would be greatly surprised if it wasn't in the several thousands.

There is no such field in the database to query, and there is no guarantee if such a field did exist that it would be accurate. The car you drive is something you type in and there is no verification to prove you actually drive that car. "Drives M3" is ambiguous and could include E30, E36, E46, and E92.

My UserCP says I drive a sprint car. Does anybody think I really drive a sprint car?

[SenorFunkyPants]

The field must exist somewhere for the vBulletin software to use to fill in everyones details. All the entries that contain some variation of E9x M3 could reasonably be considered to own/drive such a car. There will be some that may not be true just as some will contain a non E9x M3 entry such as sprint car when in fact they own an E9x M3. The number would not be far off the actual figure.

[FogCityM3]

Nice job, thx for doing this!

[regular guy]

Originally Posted by SenorFunkyPants

The field must exist somewhere for the vBulletin software to use to fill in everyones details. All the entries that contain some variation of E9x M3 could reasonably be considered to own/drive such a car. There will be some that may not be true just as some will contain a non E9x M3 entry such as sprint car when in fact they own an E9x M3. The number would not be far off the actual figure.

There is a field that you get to type in whatever you want. There's enough kids on here that say they own M3's...but don't even have their licenses. There's not any way you can depend on this for statistical correlation.

Let's say you type in:
E90
E9x
M3
E92

None of those need to mean you own an E9x M3. You could own a 335, or E30 M3.

[swamp2]

Originally Posted by Irishace

So basically an average of just around .003% of S65s fail. Of those that failed, just over 25% were supercharged and 36% of those that failed were 'modified'.

Great data here so far, now lets continue to panic about the rod bearing epidemic that plagues .003% of owners!

Not sure how you got that number. I get 0.08% (25 out of 30256). Again keep in mind this is the "m3post.com reported failure rate". The actual rate is significantly higher, no if's and's or but's.

[swamp2]

This is good work. The data thus far shows that the bearing changes that BMW made had no effect on the failure rates. The only way it doesn't show this is if the sample of folks making reports here, which admittedly is a relatively small fraction of the total M3 ownership, just happens to have a completely different group of failure rates than those who have not made a report. Statistically that is extremely unlikely.

Whether trivial or not and whether or not folks lie about what they drive I think it is a very worthy exercise to attempt to estimate how many forum members claim to drive an E9X M3. I have pm'd an admin to see if they would be willing to assist in this research. Once we have that data we can use it in an attempt to scale these reported failure rates up to a more likely actual failure rate. Yes, this will be a big approximation but it will be substantially better than what we have today.

Lastly, on a loosely related note let's compare this situation to the E46 M3 engine failure "epidemic".

E46 roadfly.com members collected 130 cases of (most if not entirely all North American) catastrophic engine failures. Total NA E46 M3 production was 43779. That is an overall reported failure rate of 0.3%. That is already almost 4 times higher. However, almost all failures were on 2001 and 2002 cars (2001 was the most dominant by far). If we use a simple straight ratio, 2 years/6 years, to capture only those two years of vehicles that gives us a much more realistic reported failure rate for those years of just North American cars. That number is 0.9%. Again, just like with this data in the OP, we must admit that only some small fraction of forum enthusiasts publicly reported their failures so the actual rate would for sure be substantially larger.

The key take-away here is that this rate is almost precisely a full order of magnitude (10x) larger (9.3 times to be specific). In short, with fairly apples to apples data (NA production and internet forum reported), this E9X M3 "problem" clearly is not an E46 M3 "epidemic" type of situation, not even close.

[regular guy]

Originally Posted by swamp2

The key take-away here is that this rate is almost precisely a full order of magnitude (10x) larger (9.3 times to be specific). In short, with fairly apples to apples data (NA production and internet forum reported), this E9X M3 "problem" clearly is not an E46 M3 "epidemic" type of situation, not even close.

That's worded weird for me. Do you mean "this rate" (E46) is larger, or do you mean "this rate" (E92) is larger? I think you mean the former, but the way it's worded is a bit confusing.

[swamp2]

Originally Posted by regular guy

That's worded weird for me. Do you mean "this rate" (E46) is larger, or do you mean "this rate" (E92) is larger? I think you mean the former, but the way it's worded is a bit confusing.

The E46 rate is larger. My wording is not as precise as it could have been, but in context as you read along and note the numbers, there can be no ambiguity.

[SenorFunkyPants]

I've attached the Excel file to the first post as a zip file if anyone wants it.

[catpat8000]

Originally Posted by swamp2

This is good work. The data thus far shows that the bearing changes that BMW made had no effect on the failure rates. The only way it doesn't show this is if the sample of folks making reports here, which admittedly is a relatively small fraction of the total M3 ownership, just happens to have a completely different group of failure rates than those who have not made a report. Statistically that is extremely unlikely.

Whether trivial or not and whether or not folks lie about what they drive I think it is a very worthy exercise to attempt to estimate how many forum members claim to drive an E9X M3. I have pm'd an admin to see if they would be willing to assist in this research. Once we have that data we can use it in an attempt to scale these reported failure rates up to a more likely actual failure rate. Yes, this will be a big approximation but it will be substantially better than what we have today.

Lastly, on a loosely related note let's compare this situation to the E46 M3 engine failure "epidemic".

E46 roadfly.com members collected 130 cases of (most if not entirely all North American) catastrophic engine failures. Total NA E46 M3 production was 43779. That is an overall reported failure rate of 0.3%. That is already almost 4 times higher. However, almost all failures were on 2001 and 2002 cars (2001 was the most dominant by far). If we use a simple straight ratio, 2 years/6 years, to capture only those two years of vehicles that gives us a much more realistic reported failure rate for those years of just North American cars. That number is 0.9%. Again, just like with this data in the OP, we must admit that only some small fraction of forum enthusiasts publicly reported their failures so the actual rate would for sure be substantially larger.

The key take-away here is that this rate is almost precisely a full order of magnitude (10x) larger (9.3 times to be specific). In short, with fairly apples to apples data (NA production and internet forum reported), this E9X M3 "problem" clearly is not an E46 M3 "epidemic" type of situation, not even close.

There really is nothing new under the sun. I made these same points in October last year, in this thread.

See the postscript here http://www.m3post.com/forums/showthr...y#post14869921

[Richbot]

Let's say we have only captured about 1/10th of the real world failures since most owners don't post here. That's still less than a 1% failure rate in the North American market.

I'm sure the publicity the E46 problem got helped to drive up forum reporting. But then again the forum was still mostly nerds back then and wasn't as mainstream as it is now. Either way I have not seen much of anything about S65 failures outside of this little world, and I heard about the E46 bearing issue back in 03 through motorsports channels. S85 yeah they have gotten some press, but mostly VANOS related, plus $100k cars get more attention when they blow up than plebe M3's

[Sti2e92sedan]

Originally Posted by Richbot

Let's say we have only captured about 1/10th of the real world failures since most owners don't post here. That's still less than a 1% failure rate in the North American market.

Not to mention the ones who already have had engine problems and replaced their engine and got rid of their car and never posted on here about their failure. Also what about the ones that have been replaced by the dealer? There isnt a known real number to that. I'm sure Bmw keeps that hidden.

[MisterEm]

Originally Posted by Sti2e92sedan

N Also what about the ones that have been replaced by the dealer? There isnt a known real number to that. I'm sure Bmw keeps that hidden.

Great point.

These failures are hidden to the best of the ability of the dealer, when possible.

Non-disclosure agreements were brought out before a repair or buyback had been agreed upon for a colleague.

2011 MY car with 30,000 miles and a "custom" rectangular hole in the block on Cylinder 5. Engine Mods: Drop in filter. Original owner, all maintenance on schedule and performed by same dealer (including filter install). Attorney he is working with now has dealt with a handful of these local cars both S85's and S65's.

He was treated so well it turned the lifelong Bimmer fan and 10 year CCA member to Audi.

You'd be awful lucky to collect the information from 1/3 of the total failures.

[SenorFunkyPants]

Originally Posted by US///M3

"Join Date: Mar 2010
Posts: 39
Mein Auto: 2010 M3 coupe
View My Garage
M3 engine failure
I leased a new 2010 M3 coupe, and after only 500 miles, I suffered a broken rod and subsequent engine failure. I opted for the DCT instead of the manual transmission.[.........]

I added this guy - I'll update the various files tomorrow.

[SenorFunkyPants]

Originally Posted by Sti2e92sedan

Not to mention the ones who already have had engine problems and replaced their engine and got rid of their car and never posted on here about their failure. Also what about the ones that have been replaced by the dealer? There isnt a known real number to that. I'm sure Bmw keeps that hidden.

Originally Posted by MisterEm

Great point.
These failures are hidden to the best of the ability of the dealer, when possible.

Its a given that not all failures are going to be reported - the best we can hope for is to extrapolate the failures reported here on M3post to give a ballpark figure for the national rate.
The fact that you guys are here shows the motivation that a unhappy ownership experience can give...I totally get that, I'm still moaning about the crappy service I got from Fuji cameras 2 years ago.
Earlier I searched Google for BMW E92 M3 engine failure and apart from the failure noted above, no previously unknown failures stood out. There just aren't that many failures to chronicle.

[US///M3]

Originally Posted by SenorFunkyPants

I added this guy - I'll update the various files tomorrow.

Thanks, you'll still get a low failure rate even if you dont include the entire production of cars for the numbers of failures on here, considering a good number of M3's are driven hard.

[swamp2]

Originally Posted by Richbot

Let's say we have only captured about 1/10th of the real world failures since most owners don't post here. That's still less than a 1% failure rate in the North American market.

I'm sure the publicity the E46 problem got helped to drive up forum reporting. But then again the forum was still mostly nerds back then and wasn't as mainstream as it is now.

Both fine points. I like to work in orders of magnitude, it's a science thing for sure. I'd guess that is also a reasonable number. We don't a 1/1 of all failures in NA reported here, and I doubt we only got 1/100 either thus 1/10 seems about right, thus 1% NA (and 1% globally) failure rate (0.8% is the actual). It's a low chance for each of us for sure but also a bit higher than I would expect from BMW, M or not, high revving or not.

[M3-S65]

Large enough sample to infer that an engine failure due to bearing problems is a low probability and even lower if you maintain your engine meticulously. A 90% probability is high enough for me and this is significantly higher. Great job OP!!

[e46e92love]

Originally Posted by US///M3

I dont think all M3 owners post here, I remember a doctor who had an engine failure that was reported on bimmerfest.
Even some who are here like the guy from Texas with the "rumad" plates ,the guy from Brooklyn, the russian guy with 3 ESS Sc'ed engine failures etc. didnt participate in the polls.
is it fair to say most dont even frequent the forums.

This story and response is probably part of why the S55 probably won't be as temperamental as the S65, because BMW probably says "we can't sell a car that requires anything other than push start and drive......special engines that require differentiated break-in periods and other user restrictions is to advanced for the masses that don't truly appreciate the difference between a 328i and an ///M3."

This is one of those failures that I take the side of BMW, which I almost NEVER DO. If the manual says one thing, and you decide you don't have to listen, in my eyes, BMW as the manufacturer has EVERY RIGHT and should hold one accountable for such carelessness. I don't care how much research there is on how to break a car in-----if you want warranty coverage you do what BMW says, not what you think.

Cheers,
e46e92

[SenorFunkyPants]

Had a PM with details of a 2008 fail to be added later today.
Author asked to be "Anon".
I'll add Filinm3s first supercharged failure as well.

Interesting chart.

[SenorFunkyPants]

Post #1 updated with new Excel entries and zip file of data.

[swamp2]

No response from the admin I tried to get some data on number of reported E9X M3 owners here. Bummer, I'll try another admin.